video-craft

Composes frame-level visual design and product demo presentation for Remotion videos.

1|Updated May 10, 2026
One-click install
npx skills add https://github.com/goheesheng/base-new --skill video-craft-goheesheng
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: video-craft
Source: https://github.com/goheesheng/base-new/tree/main/skills/launch/video-craft
Command: npx skills add https://github.com/goheesheng/base-new --skill video-craft-goheesheng

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Remotion videos often animate correctly but look generic at the frame level — flat backgrounds, floating raw screenshots, unreadable text, and weak end cards. This Skill provides composition rules, device framing patterns, and preflight checks so each frame of a product demo video looks intentionally designed. ## Core Features & Use Cases - Frame Composition Guidance: Archetypes for hero, product, split, data, and comparison frames in 16:9 and 9:16, with time-bounded readability rules and background depth treatments. - Product Demo Patterns: Remotion-ready device frames (laptop, phone, browser), cursor choreography with spring-based movement, zoom recipes, screenshot preparation, and annotation patterns. - End-Card Design: Four end-card archetypes (clean, metric, action, loop) with hold-time and CTA rules. - Use Case: You built a Remotion demo video but the frames look flat. Run this Skill to get a per-scene frame brief, wrap screenshots in device frames, add gradient-plus-noise backgrounds, and pass the preflight checklist before rendering. ## Quick Start Review my Remotion video frames for visual quality and produce a frame brief for each scene.

Frequently Asked Questions about video-craft

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make Remotion video frames look less generic?

Apply frame archetypes (hero, product, split, data, comparison) with one focal point per frame, add background depth using radial gradient plus noise or vignette, and limit text to 8 words per block. Run the preflight checklist before rendering.

How to show app screenshots in a product demo video?

Wrap screenshots in device frames — laptop, phone, or browser chrome — rather than showing raw floating images. Capture at 2x resolution, use realistic data, and add annotation callouts that stagger in after the device frame settles.

What is the difference between video-craft and marketing-video?

marketing-video handles the production pipeline: interview, storytelling, animation mechanics, and rendering. video-craft sits on top of it and governs how each static frame looks — composition, device framing, backgrounds, and end-card design.

How long should text stay on screen in a video frame?

Hold each text frame for at least (word_count x 0.3) + 1 seconds, with a maximum of 8 words per text block. If text needs more time, split it into two frames instead of extending the hold.

When should I not use this Skill?

Do not use it for requests like "create a video" or "marketing video" — those route to marketing-video, which owns the production pipeline. This Skill only fires for frame-level visual quality, composition review, and end-card design work.